Ensaios em Economia da Educação

Abstract

This dissertation comprises two independent essays in applied microeconomics within the field of economics of education. The first essay investigates the role of female peers in shaping women’s educational outcomes and teaching career trajectories in heavily male-dominated graduate programs. Using administrative data on the universe of students entering academic master’s programs in mathematics and statistics at Brazilian public universities, the study exploits quasi-random variation in cohort gender composition within programs. The results show that women in cohorts with no female peers face substantial disadvantages relative to men in degree completion, on-time completion, and dropout, and that a higher share of female peers improves all three outcomes, with no effect on men. The effect is nonlinear, concentrated in cohorts with more than 41% women, a pattern consistent with the critical mass hypothesis. Moreover, the effect is particularly pronounced in academically more demanding programs and arises only in mathematics and statistics, the sole STEM subfield in which women still complete their degrees at lower rates and drop out at higher rates than men. There is no evidence that academic scholarships or female participation in advising explain these results. Finally, cohort gender composition does not differentially affect women’s entry into or progression through teaching careers after graduation. The second essay investigates the impact of broadband internet adoption on educational outcomes in Brazilian rural public schools. Exploiting the regulatory obligations of the 2012 4G spectrum auction, which mandated the free connection of rural schools under staggered deadlines, the effects are estimated on a panel of schools from 2011 to 2018 using the multiple-period difference-in-differences estimator of Callaway e Sant’Anna (2021). The results reveal that the benefits of connectivity are concentrated in the final grades of fundamental education (grades 6–9): broadband reduces the failure rate by 0.68 percentage points and raises the approval rate by 0.61 percentage points, with additional gains in SAEB mathematics scores and in the IDEB index. Effects also grow with length of exposure, and a dose–response pattern in connection quality is found. Finally, the improvement in teacher qualification adequacy in the final grades emerges as a plausible channel for these results. Taken together, the essays offer insights for the design of educational policy on two fronts: in graduate education, policies aimed at improving gender balance in mathematics-related fields and at promoting female role models in academia, with potential spillover effects on gender gaps in mathematics at earlier educational stages; in basic education, connectivity policies that consider not only access but also the quality of the installed infrastructure, as an instrument for reducing digital exclusion and educational inequalities between rural and urban areas.